Virology 412 (2011) 366–377

Contents lists available at ScienceDirect

Virology

j ourna l homepage: www.e lsev ie r.com/ locate /yv i ro

The UL4 protein of equine herpesvirus 1 is not essential for replication orpathogenesis and inhibits gene expression controlled by viral andheterologous promoters

Robert A. Charvat, Jonathan E. Breitenbach 1, ByungChul Ahn, Yunfei Zhang, Dennis J. O'Callaghan ⁎Center for Molecular and Tumor Virology, Department of Microbiology and Immunology, Louisiana State University Health Sciences Center,1501 Kings Highway, Shreveport, LA 71130-3932, USA

⁎ Corresponding author at: Center for Molecular and TMicrobiology and Immunology, Louisiana State UniversitKings Highway, P.O. Box 33932, Shreveport, LA 711305764.

E-mail address: docall@lsuhsc.edu (D.J. O'Callaghan)1 Present address: USDA ARS, Biological Control o

(BCIRL), 1503 S. Providence Rd., Columbia, MO 65201, U

0042-6822/$ – see front matter © 2011 Elsevier Inc. Aldoi:10.1016/j.virol.2011.01.025

a b s t r a c t

a r t i c l e i n f o

Article history:Received 22 October 2010Returned to author for revision23 November 2010Accepted 19 January 2011Available online 15 February 2011

Keywords:Equine herpesvirus 1UL4 gene, EHV-1Deletion mutant, EHV-1Defective interfering particle, EHV-1Gene regulation, EHV-1

Defective interfering particles (DIP) of equine herpesvirus 1 (EHV-1) inhibit standard virus replication andmediate persistent infection. The DIP genome is comprised of only three genes: UL3, UL4, and a hybridgene composed of portions of the IR4 (EICP22) and UL5 (EICP27) genes. The hybrid gene is important forDIP interference, but the function(s) of the UL3 and UL4 genes are unknown. Here, we show that UL4 is anearly gene activated solely by the immediate early protein. The UL4 protein (UL4P) was detected at4 hours post-infection, was localized throughout the nucleus and cytoplasm, and was not present inpurified virions. EHV-1 lacking UL4P expression was infectious and displayed cell tropism and pathogenicproperties in the mouse model similar to those of parental and revertant viruses. Reporter assaysdemonstrated that the UL4P has a broad inhibitory function, suggesting a potential role in establishingand/or maintaining DIP-mediated persistent infection.

umor Virology, Department ofy Health Sciences Center, 1501-3932, USA. Fax: +1 318 675

.f Insects Research LaboratorySA.

l rights reserved.

© 2011 Elsevier Inc. All rights reserved.

Introduction

Equine herpesvirus 1 (EHV-1), a member of the Alphaherpesvirinaesubfamily, is amajorpathogenof equinesworldwide, resulting in severerespiratory, neurological, and abortigenic disease (Allen and Bryans,1986;Mettenleiter et al., 2008;O'Callaghan andOsterrieder, 2008).Viralreplication requires the gene program to be regulated in a coordinatedand temporal fashion, following a progression from immediate early(IE) to early (E) to late (L) gene expression (Caughman et al., 1985; Grayet al., 1987). EHV-1 encodes six regulatory proteins that govern the viralgene program, five being important in promoter activation and oneserving as a negative regulatory protein. The sole IE protein (IEP) trans-activates early and some late EHV-1 gene promoters and is capable oftrans-repressing its own promoter (Smith et al., 1992). Early proteinsIR4P and UL5P function synergistically with the IEP to enhanceexpression of early and late promoters (Albrecht et al., 2004; Holdenet al., 1995; Kim et al., 1997; Zhao et al., 1995). EICP0P is an earlyregulatory protein that independently serves as a potent and promis-

cuous trans-activator of EHV-1 genes of all temporal classes (Bowleset al., 1997). The late ETIF protein is a tegument protein responsible foractivating expression of the sole IE gene (Lewis et al., 1993; Purewalet al., 1994) and is essential for secondary envelopment and virus egress(von Einem et al., 2006). The early IR2 protein is a truncated form of theIE protein (Harty and O'Callaghan, 1991) and serves to down-regulatethe expression from all classes of viral promoters (Kim et al., 2006).Lastly, the EHV-1 unique IR3 gene encodes a transcript that is antisenseto the IE mRNA (Holden et al., 1992), is not translated to a detectableprotein product (Ahn et al., 2007), and functions to down-regulateexpression of the IE gene (Ahn et al., 2010).

Like many viruses (Huang and Baltimore, 1970), EHV-1 passagedat high multiplicity will form defective interfering particles (DIP) thatcan mediate a state of persistent infection (Campbell et al., 1976;Dauenhauer et al., 1982; Ebner et al., 2008; Ebner and O'Callaghan,2006). Present within the DIP genome (~7.5 kbp; Fig. 1B) are theorigin of replication, cis elements for cleavage and packaging, and onlythree genes: UL3 and UL4 conserved perfectly from the left terminusof the standard genome (~155 kbp; Fig. 1A) and a unique hybrid geneformed by a recombination event that joins portions of the IR4 andUL5 regulatory genes (Chen et al., 1996, 1999; Ebner and O'Callaghan,2006). The IR4/UL5 hybrid gene (Hyb) negatively affects expression ofmany viral genes during DIP-mediated persistent infection (Chenet al., 1999). Recent studies demonstrated that the IR4 portion of theHyb protein is important for mediating this negative regulation aswell as interfering with standard virus replication (Ebner et al., 2008).

Fig. 1. Genomes of the standard EHV-1 and EHV-1 defective interfering particles.(A) Organization of the EHV-1 standard (STD) genome and location/orientation ofthe regulatory genes and genes conserved within the DIP genome. UL unique longregion; US unique short segment; IR inverted repeat segment. Numbers show sizeof viral proteins in amino acid residues. (B) DIP genome comprised of a 7.5 kbprepeat. CPS cleavage/packaging sequence, ORI origin of DNA replication; Hybhybrid gene of IR4/UL5 sequences.

367R.A. Charvat et al. / Virology 412 (2011) 366–377

To date, no function has been reported for the UL3 or UL4 genes duringpersistent or lytic infection. In the present study, we identify theUL4 gene as an early gene and demonstrate that the UL4 protein

Fig. 2. Characterizing UL4 as an early gene through metabolic inhibitor studies, northerwith EHV-1 with a UL4 specific oligonucleotide indicates that UL4 is an early gene. (B) M(CHX) and phosphonoacetic acid (PAA) were used to treat RK13 cells infected with Eisolated from treated (+) and untreated (−) cells at 4 hpi for CHX and 8 hpi for PAA,kinase (TK), and UL4 genes. (C and D) Luciferase reporter assays were performed to exseeded in 24-well plates were transfected with 1 pmol of the pUL4–Luc reporter plasmpCMV–UL4, pCMV–ETIF and pSV–IR2; either individually (C) or in combination with thbars show standard deviation of triplicate assays.

(UL4P) has a nuclear and cytoplasmic localization and is not acomponent of purified EHV-1 virions. A possible regulatory functionof the UL4P was examined through transient transfection assayswhich revealed that the UL4P inhibited the expression of reportergenes under the control of EHV-1 promoters of all gene classes aswell as the promoters of heterologous viruses and cell genes testedto date. In addition, the absence of the UL4P resulted in an increasein the amount of viral transcripts of all gene classes duringinfection. Studies with a mutant EHV-1 lacking UL4 expressionshowed that the UL4P was not essential for viral replication in cellculture or a pathogenic phenotype in the CBA mouse model.

Results

The UL4 gene belongs to the early class in the EHV-1 gene program

A previous report described the location of the 5′ and 3′ terminiof the UL4 mRNA in relation to a TATA box and polyadenylationsignal, respectively (Harty et al., 1993). We set out to characterizeUL4 gene transcription and assign the UL4 gene to a temporal classin the EHV-1 gene program. Northern blot analysis with anucleotide probe specific for the UL4 transcript first detecteda ~0.9 kb mRNA at 2 hour post-infection (hpi) which reachedmaximal expression levels by 7 hpi (Fig. 2A). These data suggestthat UL4 is an early gene. Additionally, metabolic inhibitor studiesdemonstrated that the UL4 gene is not transcribed when proteinsynthesis is inhibited by cycloheximide (CHX; Fig. 2B); whereas, IE

n blotting, and luciferase assays. (A) Northern blot analysis of RK13 cells infectedetabolic inhibitors were used to confirm that UL4 is an early gene. Cycloheximide

HV-1 to inhibit protein synthesis or viral DNA replication, respectively. RNA wasand northern blot analysis was performed using probes for the IE, early thymidineamine the activation of the UL4 promoter by EHV-1 regulatory proteins. RK13 cellsid and 0.5 pmol of the effector plasmids pSV–IE, pSV–UL5, pSV–EICP0, pCMV–IR4,e IEP (D). Luciferase activity was measured at 48–72 hours post-transfection. Error

368 R.A. Charvat et al. / Virology 412 (2011) 366–377

mRNA, as expected, is detected in the presence and absence ofCHX. Furthermore, the UL4 transcript like that of the earlythymidine kinase (TK) transcript was synthesized when viralDNA replication was inhibited by phosphonoacetic acid (PAA).These data confirm that UL4 belongs to the early gene class, afinding further supported by the absence of a TAATGARAT motifwithin the UL4 promoter. This motif is present within the promotersof immediate early genes of other alphaherpesviruses (Lewis et al.,1997; Misra et al., 1994; Moriuchi et al., 1995), including the sole IEgene of EHV-1 (Elliott and O'Hare, 1995; Grundy et al., 1989), and isthe target sequence for binding by the viral α-trans-inducing factor(Elliott and O'Hare, 1995). Furthermore, in the reporter assaysemploying theUL4 promoter region inserted upstream of a luciferasereporter gene, the EHV-1 α-trans-inducing factor (ETIF) failed toactivate the UL4 promoter, whereas the IEP strongly trans-activatedthe UL4 promoter (Fig. 2C), confirming that UL4 is an early gene.Additional studies concerning the activation of the UL4 promoter bycombinations of plasmids that express EHV-1 regulatory proteinsindicated that the IEP alone trans-activated the UL4 promotermaximally, and that no synergistic activation occurred when theIEP was co-expressed with other EHV-1 regulatory proteins (Fig. 2D).

Characterization of the UL4 protein

To begin to characterize the UL4 protein, a rabbit polyclonal anti-UL4P specific antibody was generated (Materials and methods). Toverify the specificity of the anti-UL4P antibody, RK13 cells weretransfected with plasmids that express UL4P, UL4 fused to either thecarboxy- or amino-terminus of the green fluorescent protein (GFP), orGFP alone or cells were infected with RacL11 EHV-1. Cell lysates wereharvested and subjected to western blot analysis using the OC95 anti-UL4P antibody and a mouse monoclonal anti-GFP antibody. In thetransfected cells, the anti-GFP antibody detected bands corresponding to

Fig. 3. UL4 protein synthesis and absence of the UL4 protein in purified EHV-1 virions. Reactwestern blot analyses of lysates of RK13 cells transfected with various expression constructs48 hours post-transfection and at 24 hours post-infection were subjected to western blot anOC95 anti-UL4P antibody. (C) Western blot analyses of lysates of EHV-1 infected RK13 cellsantibodies specific for the EHV-1 UL31 DNA-binding protein, glycoprotein D, or UL4P.

GFP (26 kDa; Fig. 3A lane2), theGFP–UL4 fusion protein (46 kDa; Fig. 3Alane 7), and the UL4–GFP fusion protein (46 kDa; Fig. 3A lanes 5 and 6).This anti-GFP antibody also detected GFP (Fig. 3A lane 8) in lysates ofcells infected with the BAC-derived RacL11 virus that expresses the GFPgene inserted during the generation of the BAC (Rudolph et al., 2002).The anti-UL4P antibody detected a 23 kDa protein from cells transfectedwith UL4 expression plasmids (Fig. 3B lanes 3 and 4) as well as frominfected cell lysates (Fig. 3B lane 8). Additionally, the anti-UL4Pantibody detected theUL4–GFP fusionprotein (Fig. 3B lanes 5and6) andthe GFP–UL4 fusion protein (Fig. 3B lane 7). No bands were detectedwith the anti-UL4P antibody inmock transfected cells (Fig. 3B lane 1) orcells transfected with the GFP expression plasmid (Fig. 3B lane 2). Thus,the OC95 antibody was highly specific for the UL4P and demonstratedvery little cross reactivity with cellular proteins.

A time course experiment was completed to examine UL4protein translation. Western blot analysis of lysates of infected RK13cells with the purified rabbit anti-UL4P antibody revealed that the23 kDa protein seemed to accumulate after 4 hpi (Fig. 3C). Todetermine whether the UL4 protein is a component of the EHV-1virion, purified virus particles were prepared and subjected towestern blot analyses. Nitrocellulose membranes were reacted withprimary antibodies specific for the UL4P, glycoprotein D (gD), anenvelope glycoprotein essential for EHV-1 replication (Csellneret al., 2000; Flowers and O'Callaghan, 1992; Frampton et al.,2005), and the EHV-1 nonstructural UL31 protein, the major DNAbinding protein (Lewis et al., 1995). As shown in Fig. 3D, thepreparation of virions was free of cellular protein contamination asdemonstrated by the lack of the nonstructural UL31 protein in thevirion preparation, whereas this protein was present in large amountsin lysates of infected cells. As a positive control, the anti-gD antibodyreacted with viral protein present in both infected cell lysates andpurified virions. In contrast, the anti-UL4P antibody reacted stronglyonly with infected cell lysates and failed to react with proteins in the

ivity and specificity of the OC95 anti-UL4 protein polyclonal antiserum were verified byand lysates of cells infected with BAC-derived (bd) RacL11 EHV-1. Lysates prepared atalysis and reacted with either (A) a mouse monoclonal antibody to GFP or (B) purifiedwith the purified anti-UL4P antibody. (D) Western blot analysis of purified virions with

369R.A. Charvat et al. / Virology 412 (2011) 366–377

purified virion preparation, indicating that the UL4P is not acomponent of the EHV-1 virion.

Utilizing themonospecific anti-UL4P antibody, immunofluorescenceassays were completed to determine the cellular localization of theUL4 protein in RK13 cells transfected with plasmids expressingeither the UL4 gene or the UL4–GFP fusion gene, or cells infected withEHV-1. Expression of the UL4 protein (Fig. 4A) or the UL4–GFPfusion protein (Fig. 4B) in the absence of other viral proteinsresulted in a broad distribution of the UL4P within both thecytoplasm and nucleus of these cells. A similar fluorescent stainingpattern was observed in cells infected with EHV-1 (Fig. 4C),suggesting that the 23 kDa UL4 protein localizes within both thenucleus and cytoplasm during infection. To verify the localizationof the UL4 protein in both cellular compartments, RK13 cells wereinfected with EHV-1, and cell lysates were separated intocytoplasmic and nuclear fractions as described in the Materialsand Methods. Western blot analysis of the proteins isolated fromthese cellular fractions confirmed that the UL4 protein was locatedwithin both the cytoplasm and nucleus (data not shown).

UL4P is an inhibitor of gene expression and antagonizes IEPtrans-activation of EHV-1 promoters

Although known to be conserved perfectly within the genome ofdefective interfering particles (Baumann et al., 1984, 1986; Ebneret al., 2008; Ebner and O'Callaghan, 2006), no function has beenreported for UL4. Herpes simplex virus 1 (HSV-1) UL55, the UL4Phomolog, was demonstrated to inhibit transient gene expressionwhen co-transfected into cells along with plasmids expressing theimmediate early gene α27 as well as alpha genes α0 and α4 (Blocket al., 1991). Additionally, Yamada et al. (1998) characterized theUL55 protein of HSV-2 as a potential accessory protein in virionassembly and maturation and found it to co-localize with capsidprotein ICP35 at the periphery of infected cell nuclei. To determinewhether UL4P possesses any regulatory activity, luciferase reporterassays were performed. The reporter plasmid expressed the fireflyluciferase gene under the control of various EHV-1 promoters,including the immediate early (IE), early promoters EICP0, thymidinekinase (TK), and IR2, and late promoters ETIF and glycoprotein K (gK).

Fig. 4. Localization of the UL4 protein in transfected and EHV-1 infected cells by immunofluo(A) pCMV–UL4 or the (B) pUL4–GFP expression plasmid, or (C) were infected with RacL1112 hours post-infection and permeabilized with 0.2% Triton X100 in PBS. Panels A and C: puranti-rabbit IgGwas the secondary antibody. Panel B shows fluorescence due to GFP expressiopanel). Arrows in the upper panels indicate the nuclei of the cells of interest stained in the

Rabbit kidney cells were transfected with the various reporterplasmids along with the effector plasmids pSV–IE and/or pCMV–UL4,and luciferase activity was measured at 48–72 hours post-transfection.For the IE promoter, luciferase activity was inhibited by 77% in thepresence of the UL4 protein (Fig. 5A). In a separate reporter assayemploying the chloramphenicol acetyltransferase (CAT) gene underthe control of the EHV-1 IE promoter, UL4 protein expressioninhibited CAT activity by 87% (data not shown). UL4P expressiondecreased luciferase activity for the early EHV-1 TK and IR2promoters 95% and 85%, respectively (Figs. 5C and D). In addition,luciferase activity driven by EHV-1 late promoters was negativelyaffected by the UL4P, such that in the case of the ETIF and gKpromoters the inhibition reached levels of 78% and 98%, respectively(Figs. 5E and F). Conversely, when cells were transfected with the IEeffector plasmid, luciferase activity was greatly increased for the earlypromoters and theETIFpromoter;whereas, the late gKpromoter and theIE promoter were inhibited by expression of the IE protein, confirmingour previous reports (Caughman et al., 1985; Harty and O'Callaghan,1991; Kim et al., 1999; Smith et al., 1992, 1994). When the UL4 and IEexpression plasmids were co-transfected into cells, the ability of the IEPto trans-activate the early EICP0, TK, IR2, and late ETIF promoters wasdiminished (Figs. 5B to E). IEP trans-activation of these fourpromoters was inhibited by 45% to 70%. The IEP has been shown toinhibit its own promoter (Harty and O'Callaghan, 1991; Smith et al.,1994), and the coexpression of the UL4 and IE proteins furtherabrogated the luciferase activity driven by the IE promoter (Fig. 5A).These data suggest that the UL4 protein plays an inhibitory role ingoverning the expression of EHV-1 genes.

Given that the UL4P appeared to function across all gene classes,we next wanted to establish if the negative effect on reporterexpression was specific for EHV-1 genes. In a similar reporter assay,luciferase activity was measured from reporter plasmids that usedthe simian vacuolating virus 40 (SV40) large T antigen promoterand the human cytomegalovirus (CMV) IE promoter. In thepresence of UL4P, luciferase activity was decreased 90% fromthese heterologous viral promoters as well (Fig. 6A), indicatingthat UL4P inhibition is not virus specific. The inhibitory effect of theUL4P was not specific for the firefly luciferase reporter gene assimilar levels of inhibition were obtained in assays that employed

rescence with purified anti-UL4P antibody. RK13 cells were transfected with either theEHV-1. Cells were fixed with 4% paraformaldehyde at 48 hours post-transfection or atified anti-UL4P antibody was used as the primary antibody and an Alexa Fluor 488 goatn. Coverslips were mountedwith a DAPI solution to stain nuclei (top photograph of eachlower panels.

Fig. 5. Luciferase assays examining the effect of the UL4 protein on luciferase expression driven by EHV-1 promoters of all gene classes. RK13 cells in 24-well plates were transfectedwith 0.5 pmol of the expression plasmids pCMV–UL4, pSV–IE, or both plasmids, along with 1 pmol of the luciferase reporter plasmid under the control of the EHV-1 (A) immediateearly promoter; or early promoters (B) EICP0, (C) thymidine kinase (TK), and (D) IR2, or late promoters (E) ETIF and (F) glycoprotein K (gK). Luciferase activity was assayed at 48 hourspost-transfection as described in Materials and methods.

370 R.A. Charvat et al. / Virology 412 (2011) 366–377

the renilla gene as the reporter (data not shown). Lastly, a numberof cellular promoters was tested to ascertain whether expression ofthe UL4 protein affects non-viral gene expression. To this end,expression of the UL4 protein was found to inhibit luciferaseactivity driven by the GAPDH, eIF4E, IFNβ, and HPRT cellularpromoters by 44%, 86%, 75%, and 78%, respectively (Fig. 6B). Thesedecreases in luciferase activity were not attributable to cytotoxicityin the presence of the UL4 protein, as demonstrated by similarlevels of cell viability in cells expressing the UL4P compared to non-transfected cells (data not shown). Taken together, the datapresented in Figs. 5 and 6 indicate that the UL4 protein negativelyaffects gene expression, not only from viral promoters but cellularpromoters as well.

UL4 is not required for EHV-1 lytic replication

After observing the inhibitory activity of the UL4 protein against abroad range of viral and cellular promoters, we wanted to examine ifUL4 plays an essential role in EHV-1 lytic infection. Itwas hypothesized

that deletion of the UL4 gene would increase the ability of the virusto replicate due to the absence of the additional inhibitory activityof the UL4P. To begin to address this hypothesis, a UL4-null EHV-1was generated utilizing the galK bacterial artificial chromosome(BAC) technology (Rudolph et al., 2002; Warming et al., 2005).Initially, the UL4 gene was replaced by the galK gene to create aUL4-deleted virus. The galK gene was then replaced with a mutantform of UL4 that contained a STOP codon at position 18 within theamino acid sequence (UL4aa18stop) or a wild-type UL4 to generatea revertant virus (ΔUL4R). Once the mutant and revertant BACswere generated, the gross genomic structure was examined byBamH I restriction enzyme digestion. Fig. 7A demonstrates that theprocess used for generating the mutant and revertant BACs did notresult in any spurious recombination events or alterations in theoverall genomic structure, as evidenced by similar digestionpatterns of the parental RacL11, ΔUL4R, and UL4aa18stop BACs.The second BamH I fragment in the ΔUL4 (galK+) BAC migrated inagarose gels more slowly than the other BACs, indicative of theadditional base pairs within the galK gene compared to the UL4

Fig. 6. Luciferase assays examining the effect of the UL4 protein on luciferase expressiondriven by heterologous viral promoters or cellular promoters. (A). Luciferase activitydriven by the SV40 large T antigen promoter or the CMV IE promoter as comparedto activity controlled by the EHV-1 IE promoter. (B). Luciferase activity driven bycellular promoters glyceraldehyde phosphate dehydrogenase (GAPDH), eukaryoticinitiation factor 4E (eIF4E), interferon β (IFNβ), or hypoxanthine–guaninephosphoribosyltransferase (HPRT).

Fig. 7. Generation and verification of the UL4aa18stop EHV-1 and ΔUL4R revertantEHV-1 by galK BAC technology. Homologous recombination was used to delete the UL4gene as described in Materials and methods. Briefly, PCR was utilized to append UL4DNA flanking sequences to the galK gene. SW106 E. coli harboring the EHV-1 BAC weretransformed with the resultant PCR product and plated on positive selection plates. Togenerate the revertant BAC, wild-type UL4 sequences were transformed into bacteriacontaining the mutant BAC and plated on negative selection plates with 2-deoxygalactose. The mutant and revertant BAC genomes were analyzed and verifiedby (A) BamH I restriction enzyme digestion and (B) Southern blot analysis. (C) RK13cells were transfected with each of the three BAC genomes to generate wild-type,UL4aa18stop, and ΔUL4R viruses. RK13 cells were then infected with each virus, and at9 hours post-infection lysates were prepared and subjected to western blot analysiswith the purified anti-UL4P antibody.

371R.A. Charvat et al. / Virology 412 (2011) 366–377

gene (Fig. 7A; lane 2). Southern blot analysis for the presence of theUL4 gene within the various BACs revealed that the UL4 sequencewas absent from the ΔUL4 (galK+) BAC, while the UL4-specificprobe hybridized to DNA sequences within the parent, mutant, andrevertant BACs (Fig. 7B). Furthermore, PCR analysis and DNAsequencing verified that the BACs were correct (data not shown).

To determine whether UL4 was essential for EHV-1 replication,the BACs were transfected into permissive RK13 cells. As shown inFig. 8A, each BAC resulted in the formation of plaques in transfectedcells, indicating that UL4 is dispensable for EHV-1 lytic replication.However, it was observed that the plaques formed by theUL4aa18stop virus were smaller in size compared to those of theparent and revertant viruses, being approximately one-third the sizeof wild-type and revertant plaques (Fig. 8B). To verify that the stopmutant virus did not produce UL4 protein, lysates from cellsinfected with the wild-type, mutant, and revertant viruses weresubjected to western blot analysis. As expected, the UL4 protein wasdetected in cells infected with the wild-type RacL11 and ΔUL4Rviruses, but not in cells infected with the UL4aa18stop virus(Fig. 7C). These data indicated that the UL4 protein is not essentialfor EHV-1 replication, but the smaller plaque size of the UL4aa18-stop virus may indicate a deficiency in its ability to replicate orspread efficiently.

Single-step growth kinetics experiments performed to addressthis possibility revealed that the UL4aa18stop mutant replicatedwith similar kinetics and to equivalent levels as the wild-type andrevertant viruses that express UL4 (Fig. 9A). Similar virus titers wereobtained from the extracellular supernatant of RK13 cells infectedwith all three viruses, indicating that impairment of virion release isnot responsible for the smaller plaque phenotype. Finally, cell

tropism was examined to determine whether UL4P plays a role inexpanding the host range of EHV-1, as shown for the EHV-1 earlyIR4 regulatory gene (Breitenbach et al., 2009) as well as for the IR4homologue EICP22 of HSV-1 (Poffenberger et al., 1993; Post andRoizman, 1981). The UL4aa18stop virus was able to replicate in cellsof mouse, rabbit, equine, monkey, and human origin (Fig. 9B). Virustiters in cells infected with all three viruses were similar, withexceptions of mouse L-M and human HeLa cells, for which anapproximately one log reduction in maximal virus titer wasobserved for the UL4aa18stop mutant virus, indicating that UL4does not contribute to the broad host range of EHV-1 (O'Callaghanand Osterrieder, 2008). However, upon detailed statistical analysisof the viral titers from multiple experiments, we found that thevalues were significantly different between wild-type and theUL4aa18stop EHV-1 for all cell types. Although the levels ofmaximal virus production were statistically different, the biologicalrelevance is not likely to be consequential.

The UL4 gene is not a virulence factor in the CBA mouse model ofEHV-1 pathogenesis

Although no striking differences were observed between theUL4-mutant virus and the parental EHV-1 in their abilities to infectand replicate in cell culture, we wanted to determine whether theUL4 gene plays a role in EHV-1 pathogenesis. Utilizing the wellcharacterized CBA mouse model of EHV-1 infection (Colle et al.,1996; Frampton et al., 2002, 2004; Matsumura et al., 1996;O'Callaghan and Osterrieder, 2008; Smith et al., 2000, 2005), groupsof 12 female mice were intranasally inoculated with wild-type,UL4aa18stop, or ΔUL4R viruses as well as with sterile medium for

Fig. 8. Morphology of plaques produced on RK13 cells transfected with the parent, UL4aa18stop, and ΔUL4R BACs. (A) RK13 cells were transfected with the various BACs, overlaidwith medium containing 1.5% methylcellulose, and examined after 5 days for plaque formation by staining with 0.5% methylene blue dye. (B) Plaque size was analyzed with theImageJ software (Materials and Methods). 30 plaques produced by each virus were measured and the Student's-t test was used for statistical analysis.

Fig. 9. Growth kinetics and cell tropism of the wild-type EHV-1, UL4aa18stop EHV-1 andΔUL4R EHV-1. (A) RK13 cells in 60-mm dishes were infected with the wild-type, mutant,and revertant viruses at an MOI of 5. Virus was harvested from the extracellularsupernatant as well as from freeze–thawed cells at the indicated times post-infection andenumeratedbyplaqueassay. (B) Theability of thewild-type,mutant, and revertant virusesto replicate in various cell types was determined by infecting mouse L–M, rabbit RK13,equine NBL-6, African green monkey Vero, and human HeLa cells at a MOI of 5. Maximalviral titers at 72 hours post-infection were determined by plaque assay on RK13 cells.

372 R.A. Charvat et al. / Virology 412 (2011) 366–377

a control group, and monitored for signs of morbidity andmortality over the course of 9 days. As expected, mice in thesterile medium group demonstrated no signs of infection andexhibited increases in body weight (Fig. 10A). On the other hand,all mice infected with each of the three EHV-1 displayed ruffledfur and a huddling phenotype, along with significant decreases inoverall body weight (Fig. 10A). EHV-1 lacking UL4 gene expressioncaused decreases in body weight similar to that of the wild-typehighly pathogenic EHV-1, suggesting that UL4 is not required forEHV-1 pathogenesis in the mouse. Furthermore, mortality rateswere similar for mice infected with either the wild-type virus orthe UL4aa18stop EHV-1 (Fig. 10B). In cell culture, there was anapproximate one log decrease in the ability of the UL4aa18stopmutant virus to replicate in mouse L–M cells compared to theparental virus (Fig. 9B). To determine whether the reduced abilityof the UL4-mutant virus to replicate in mouse cells in culture waslikewise reflected in vivo, infected mice were sacrificed at days2, 3, and 4 post-infection, and virus was titered from infectedlungs. As shown in Fig. 10C, the UL4aa18stop virus replicated asefficiently as the wild-type EHV-1 in murine lungs. Taken together,the results from these animal experiments indicate that the UL4gene is not a determinant of virulence for EHV-1.

Viral transcripts of all gene classes are increased in the absence of UL4protein synthesis

We previously observed that the UL4P was capable of inhibitingluciferase gene expression driven by representative EHV-1 promotersfrom all three gene classes (Fig. 5). To extend these findings, the effectof UL4P on the synthesis of EHV-1 transcripts was addressed. Cellswere infected with wild-type virus or the UL4aa18stop EHV-1 andexamined for the effect of the absence of the UL4 protein synthesis onthe level of selected IE, E, and L transcripts. RNA was isolated andanalyzed by northern blot assays at 4 hpi for IE, 8 hpi for early TK, and12 hpi for late gK gene expression. As shown in Fig. 11, the levels of

Fig. 10. Comparison of the pathogenic properties of wild-type, UL4-mutant and revertantEHV-1 in the CBA mouse model. Groups (n=12 mice) of 6 week old female CBA micewere mock infected with sterile medium or were infected intranasally with 1.25×106pfuof wild-type RacL11, UL4aa18stop, or ΔUL4R EHV-1. Mice were monitored daily forclinical signs of infection, gain or loss in body weight, and morbidity. (A) Changes in bodyweight; (B) Percent survival; and (C) Virus lung titer. Whole lungs were harvested frommice sacrificed on days 2, 3 and 4 post-infection, and virus was titered by plaque assay.

Fig. 11. Northern blot analysis examining gene expression in cells infected with wild-type (WT) or STOP-mutant EHV-1. RK13 cells were infected with wild-type EHV-1 orthe UL4aa18stop EHV-1, and RNA was isolated at 4, 8, and 12 hours post-infection.Northern blot analysis was used to determine transcript levels for the immediate early(IE) gene, the early thymidine kinase (TK) gene, and the late glycoprotein K (gK) gene.

373R.A. Charvat et al. / Virology 412 (2011) 366–377

the transcripts for each of the three viral genes were greater in theabsence of the UL4 protein as compared to those of WT infected cells.Densitometry measurements revealed that the levels of the IE, TK, andgK transcripts were increased a minimum of 20% in the absence of theUL4P, and in repeated experiments these increases were as great as42%. These data indicate that the absence of the UL4P allows forincreased viral transcription at all times during infection, and that theUL4P is indeed an inhibitor of EHV-1 gene expression.

Discussion

The EHV-1 UL4 gene, conserved within the genome of defectiveinterfering particles, has no attributed function during lyticreplication or DIP-mediated persistent infection (Baumann et al.,1984, 1986; Ebner et al., 2008; Ebner and O'Callaghan, 2006). In thisreport, we presented findings suggesting that this early geneencodes an inhibitory protein that negatively affects viral andcellular gene expression. The expression of the UL4 protein wascapable of inhibiting gene expression driven by promoters from allEHV-1 gene classes by greater than 75%. These data appear verysimilar to the results from experiments concerning the IR2 protein, anegative regulatory protein that was demonstrated to down-regulate all EHV-1 gene promoters tested to date (Kim et al.,2006). Comparing the broad negative regulatory activity of the IR2Pand UL4P suggests that UL4P may serve as an auxiliary inhibitoryprotein to govern EHV-1 gene expression during lytic infection.However, this inhibitory function appears to differ from thatreported for the HSV-1 UL55 protein (Block et al., 1991), whichshares only 32% amino acid identity with UL4P. In that report, theauthors concluded that the inhibition of transient gene expressionrelied on the presence of other alpha genes expressed in conjunctionwith the UL55 protein, such that the UL55 protein alone was notsufficient to negatively affect viral gene expression. Conversely, in thecase of the UL4 protein, EHV-1 gene expression was significantlyinhibited when only the UL4P was expressed. Although these homologsmay have an inhibitory function in common, our findings suggest thattheir mechanism of action is different. However, we cannot rule out thepossibility that in the prior study involving HSV-1 UL55, alpha geneswere required solely for the efficient expression of the UL55 gene. Inthat case, the alpha genes per se contributed no inhibitory function, andthus it is conceivable that UL4 and UL55 proteins function through asimilar mechanism. We are interested in exploring how the UL4protein functions to inhibit gene expression, and current work isfocused on elucidating a mechanism. Due to the broad inhibitoryactivity of the UL4P, it appears that the mechanism is not gene-specific as supported by the fact that the UL4P does not bind tospecific promoter DNA sequences (data not shown). We haveconsidered the possibility that the UL4 protein interacts withanother protein involved in gene expression. GST-pulldown assaysare being utilized to examine protein-protein interactions that mayoccur between UL4P and other viral and/or cellular proteins. Inpreliminary studies, the UL4 protein did not interact with the IEP;however, it appears that the UL4P functions at the level of genetranscription via an interaction with general transcription factors(unpublished data).

To determine whether the inhibitory function of UL4P wasimportant for viral replication, a mutant virus (UL4aa18stop) wasgenerated. Surprisingly, we found that the replication kinetics, celltropism, and pathogenesis of an EHV-1 lacking UL4 expression weresimilar to those of the parental and UL4 revertant viruses (Figs. 9and 10). These findings were contrary to our hypothesis thatremoving UL4 gene expression would allow enhanced growth of themutant virus compared to wild-type EHV-1. As UL4 is not a diploidgene, serial passage of the UL4aa18stop EHV-1 was not expected toresult in its reversion to a wild-type phenotype. Indeed, theUL4aa18stop virus failed to produce the UL4 protein upon serial

374 R.A. Charvat et al. / Virology 412 (2011) 366–377

passage in cell culture (data not shown). The varicella-zoster virus(VZV) ORF3 protein is a UL4P counterpart that has 37% identity inamino acid sequence. Similar to these findings with the UL4-mutantEHV-1, it was shown that deletion of VZV ORF3 did not affect virusreplication in vivo or in vitro (Zhang et al., 2007). One significantdifference between the UL4aa18stop EHV-1 and the parental viruswas a reduction in plaque size (Fig. 8). One possible explanation forthe smaller plaques is that the virus has a defect in the ability toegress from infected cells, but that possibility was not supported bythe observation that the extracellular virus titers were identical forRK13 cells infected with the parental virus and the UL4aa18stopvirus.

During DIP-mediated persistent infection, it is believed that thehighly expressed IR4/UL5 hybrid protein is responsible for mediat-ing the interference of standard virus replication, as demonstratedby the observation that recombinant DI particles deleted of the UL3and UL4 genes and engineered to express only the hybrid proteinretained DIP interference activity (Ebner et al., 2008). However, itremains unclear what actually triggers the establishment ofpersistent infection and how high levels of DIP maintain persistentinfection. Here, we presented data that demonstrate that the UL4protein negatively affects all EHV-1 gene classes. As this gene isconserved perfectly within the DIP genome and expressed at veryhigh levels in persistently infected cells (Gray et al., 1989), it ispossible the UL4 protein could greatly impair the expression ofstandard viral genes. This possibility is supported, in part, by thefinding that EHV-1 gene expression is tempered in cells infectedwith the wild-type virus that expresses UL4P, as compared to geneexpression in cells infected with the UL4aa18stop mutant EHV-1. Thisbroad inhibitory activity may facilitate the establishment and/ormaintenance of DIP-mediated persistent infection. As attractive as thismay be, the studies to determine the role UL4 plays during persistencywill require a better understanding of the function(s)mediated by thisearly gene product.

These studies are the first to characterize UL4 as a member ofthe early class within the EHV-1 gene program. The UL4 proteinlocalizes within both the nucleus and cytoplasm of infected cells,but is not a component of mature virus particles. The luciferasereporter assays demonstrated that the UL4P is an inhibitory protein,a finding that was supported by the results of the viral geneexpression studies using the UL4aa18stop EHV-1. Thus, the UL4protein joins the IR2P as the second EHV-1 early gene product thatpossess predominantly negative activity. Current studies areaddressing how this protein mediates this inhibitory outcome inthe case of EHV-1 genes and if this mechanism applies to genes ofheterologous viruses and host cells.

Materials and methods

Cell culture and virus propagation

Mouse L-M, rabbit RK13, equine NBL-6, African green monkeyVero and human HeLa cells were grown in Dulbecco's minimumessential medium (DMEM) supplemented with either 5% or 10% fetalbovine serum at 37°C in a 5% CO2 incubator. The pathogenic EHV-1strain RacL11 was propagated in equine NBL-6 cells, while theattenuated KyA strain was grown in mouse L–M cells. After growing ahigh-titer stock of KyA virus, the supernatant virus was collected andclarified by removing cells by centrifugation. Highly purified virionswere prepared as previously described in detail (O'Callaghan andRandall, 1976; Perdue et al., 1974).

Plasmids

Manipulation and cloning of the expression plasmids wereperformed according to the procedures described by Sambrook et al.

(1989). pGEX-UL4, which expresses the UL4 gene as a GST fusionprotein, was generated by PCR amplifying the UL4 gene from theRacL11 BAC (Rudolph et al., 2002) to include the restriction enzymesites EcoR I and Sal I with forward primer: 5′-CATGAATTCC-CATGCTGCCGGCAAACCGCGCAGAACAC-3′ and reverse primer: 5′-CTAGGTCGACTTATCGTTTATTTTCTCGCTGGCGCTCTTTGGCCGA-3′. ThePCR product and the pGEX-4T-2 plasmid were digested with theappropriate enzymes and ligated. pGFP-UL4 expresses a GFP–UL4fusion protein and was generated by PCR using forward primer: 5′-CATGTGTACAAGATGCTGCCGGCAAACCGCGCAGAA-3′ and reverseprimer: 5′-CATTGCGGCCGCTTATCGTTTATTTTCTCGCTG-3′ to appendrestriction enzyme sites BsrG I and Not I, respectively, to the UL4 gene.The digested insert was cloned into the pEGFP–N1 plasmid thatcontains the enhanced GFP gene under the control of the HCMV IEpromoter. The pUL4–GFP plasmid expresses an amino-terminalUL4–GFPfusionprotein.PCRwasusedtoattachEcoRIandBamHIenzymesites to the UL4 gene using forward primer: 5′-AATTGAATTCCGCCAC-CATGCTGCCGGCAAACCGCGCAGAAC-3′ and reverse primer: 5′-CATGG-GATCCCGTTTATTTTCTCGCTGGCGCTCTTTG-3′, respectively. Thedigested fragment was again inserted into the pEGFP–N1 plasmid.To generate pCMV–UL4, an expression plasmid that expresses theUL4 gene under the control of the HCMV IE promoter, forwardprimer: 5′-AATTGAATTCCGCCACCATGCTGCCGGCAAACCGCGCA-GAAC-3′ and reverse primer: 5′-CATTGCGGCCGCTTATCGTT-TATTTTCTCGCTG-3′ were used to add the EcoR I and Not I sites tothe UL4 gene, respectively. The digested insert was then cloned into.1pt?>the pEGFP–N1 plasmid to replace the GFP gene. pUL4–Luc, areporter plasmid that contains the firefly luciferase gene driven bythe UL4 promoter, was created using PCR to attach Kpn I and Bgl IIrestriction sites flanking the UL4 promoter region. Forward primer:5′-CATGGTACCCCAACGCAAACAGTTGGCACCGTG-3′ and reverseprimer: 5′-CATAGATCTCAGGCTGGGAATTTGCTCGACTGAAG-3′ wereused. The resulting 1.5 kb enzyme fragment encompassing the UL4promoter with TATA box was inserted into the pGL3–Basic plasmid.The remaining expression and reporter plasmids were generatedelsewhere (Bowles et al., 2000; Holden et al., 1995; Kim andO'Callaghan, 2001; Smith et al., 1992; Zhao et al., 1995).

Expression and purification of GST–UL4 fusion protein

The induction of GST fusion protein synthesis and its purificationhave been described previously (Albrecht et al., 2004; Jang et al., 2001).To prepare a purified UL4 protein lacking the GSTmoiety, GST–UL4wastreated with a thrombin cleavage capture kit (Novagen, Madison, WI)per the manufacturer's instructions, and the GST portion wasremoved using GST-bind resin beads (Novagen). The purified proteinwas concentrated using Amicon Ultra centrifugation filter deviceswith a size exclusion of 10 and 30 kDa (Millipore, Billerica, MA).

Generation of anti-UL4 protein polyclonal antibody

Antibody generation has been described elsewhere (Albrecht et al.,2004). Briefly, two New Zealand White rabbits (approximately 4 kgper rabbit) were immunized with either the GST–UL4 fusion protein(OC94) or the purified UL4 protein (OC95). The primary inoculumwasemulsified in Freund's complete adjuvant, and after 4–8 weeks,booster immunizations emulsified in incomplete Freund's adjuvantwere administered every 14 days for a total of five booster injectionsfor each rabbit. Before each booster injection, small quantities ofserum were drawn from each rabbit to test antibody titers before thefinal bleed was performed. The anti-UL4P antibody was purified usingprotein A agarose beads (Pierce, Rockford, IL) according to themanufacturer's directions. The working dilution in western blotassays was 1:5000 or greater.

375R.A. Charvat et al. / Virology 412 (2011) 366–377

Immunofluorescence microscopy and cell fractionation

RK13 cells were seeded onto No. 1.5 glass coverslips (FischerScientific, Pittsburgh, PA) in the bottom of a 6-well plate. The cellswere either infected with wild-type RacL11 EHV-1 or transfectedwith the expression plasmids pCMV–UL4 or pUL4–GFP. Infectedcells were fixed at 12 hours post-infection, while transfected cellswere fixed at 48 hours post-transfection. Cells were fixed with a 4%paraformaldehyde solution in PBS, and cellular membranes werepermeabilized with 0.2% Triton X100 in PBS. After washing the cellswith PBS, 10% goat serum in PBS was used to block non-specificantibody interactions. Anti-UL4P OC95 antiserum was used as theprimary antibody at a dilution of 1:100 in PBS/0.1% Tween-20 andthe secondary antibody was Alexa Fluor 488 goat anti-rabbit IgG(Invitrogen, Carlsbad, CA) at 1:100 in PBS/0.1% Tween-20. Unboundantibody was removed by washing, and the coverslips wereremoved from the 6-well plates and placed on glass slides. SlowFadeGold antifade reagent with DAPI (Invitrogen) was used as themounting solution (arrows indicate nuclei in Fig. 4). Clear nail polishsealed the edges of the coverslips, and the slides were viewed on aNikon Eclipse TE300 inverted fluorescent microscope (Melville, NY).Cells were separated into cytoplasmic and nuclear fractions by thefollowing procedure, modified from Muramatsu et al. (1963). Cellswere washed with cold PBS, then scraped and pelletted in amicrofuge tube. Resuspend pellets in ice-cold Buffer A (10 mMHEPES, pH 7.9; 1.5 mMMgCl2; 10 mMKCl; 0.5 mMDTT) and keep onice for 5 min. Cells were disrupted by repeated passage through a 26gage needle. Disrupted cells were centrifuged at 300 g for 5 min at4 °C to pellet the nuclei. The supernatant was saved as thecytoplasmic fraction. The nuclear pellet was resuspended in sucrosebuffer A (0.25 M sucrose, 10 mM MgCl2), layered over a cushion ofsucrose buffer B (0.88 M sucrose, 0.5 mM MgCl2), and centrifuged at3000 g for 10 min at 4 °C. Nuclear pellets were disrupted with RIPAbuffer (50 mM Tris, pH 7.5; 150 mM NaCl; 1% NP-40; 0.5%deoxycholate) and centrifuged at 3000 g for 10 min at 4 °C to pelletinsoluble debris.

Luciferase reporter assays

RK13 cells were seeded into 24-well plates and used thesubsequent day at a confluency of 80%. Cells were transfectedusing lipofectin (Invitrogen) and Opti-MEM medium (Gibco, BRL,Carlsbad, CA) as detailed elsewhere (Ahn et al., 2007). Briefly, 1 pmolof the various reporter plasmids and 0.5 pmol of the expressionplasmids were co-transfected into RK13 cells. Eight hours later,regular growth medium was added, and luciferase activity wasassayed at 48–72 hours post-transfection utilizing a luciferase assaykit (Promega, Madison, WI) and a POLARstar OPTIMA plate reader(BMG LABTECH Inc., Cary, NC) following the manufacturer'sinstructions.

Generation and confirmation of UL4-null EHV-1

A RacL11 BAC deleted for the UL4 genewas generated through galKBAC recombineering (Ahn et al., 2010; Warming et al., 2005). Toreplace the UL4 gene, PCR amplification (forward primer: 5′-TTATCGTTTATTTTCTCGCTGGCGCTCTTTGGCCGAGGTTATTCCCC-TAGCCTGTTGACAATTAATCATCGGCA-3′ and reverse primer: 5′-ATGCTGCCGGCAAACCGCGCAGAACACTCATCTGATGCAGAGCCGCGG-GATCAGCACTGTCCTGCTCCTT-3′) was used to append UL4 flankingsequences to the ends of the galK selection gene, and the subsequentPCR product was transformed into SW106 E. coli harboring the RacL11BAC. Bacteria were plated onto positive selection agar plates contain-ing galactose, and colonies were screened by PCR (5′-CAGACCCA-GAGCTCCACGCACCGTCC-3′/5′-GCAGATCTTGCTCCCAGACCTGACC-3′and 5′-CCCTCTTCTCGAACACGCCGATGAAAAAGGCG-3′/5′-GGCAGA-

TACCTGCAGCCTTGTATCGGCC-3′) for the correct junction sequences.The galK marker was replaced with a mutant form of the UL4 genethat contains a STOP codon at position 18 (underlined; UL4aa18-stop) or a wild-type form of the UL4 gene to generate a UL4revertant BAC (ΔUL4R). PCR was used to create the desired DNAfragments that were transformed into the E. coli harboring thegalK inserted RacL11 BAC (5′-TTATCGTTTATTTTCTCGCTGGCGC-3′/5 ′ -ATGCTGCCGGCAAACCGCGCAGAACACTCATCTGATGCA-GAGCCGCGGGATTAAATAGGCTCCCACGGGAGGA-3′ and 5′-TTATCGTTTATTTTCTCGCTGGCGCTCTTTGGCCGAGGTTATTCCCCTAG-3′/5′-ATGCTGCCGGCAAACCGCGCAGAACACTCATCTGATGCA-GAGCCGCGGGA-3′). Transformed bacteria were plated on negativeselection plates containing glycerol and 2-deoxygalactose. Result-ing colonies were again screened by PCR for the junctionsequences as well as DNA sequence analysis. The desired BACswere further verified by BamH I digestion and Southern blotanalysis. To reconstitute the mutant and revertant RacL11 BACs asstandard virus, the BAC sequences were replaced with viral DNAsequences for the EUs4 gene by transfecting RK13 cells with bothBAC and a plasmid containing the EUs4 sequences. Plaques lackingGFP expression were plaque purified for three rounds and thenpropagated to high titer on equine NBL-6 cells.

Southern, northern and western blot analyses

Mutant and revertant RacL11 BACs were subjected to BamH Idigestion and then electrophoretically separated on a 0.6% agarosegel. Digested DNA was transferred onto a positively-charged nylonmembrane (Ambion, Austin, TX) using a semi-dry electroblotter(Bio-Rad Laboratories, Hercules, CA). The transferred DNA washybridized with a fragment of the UL4 gene (PCR product forwardprimer: 5′-GGCCTGGGCAGAGTTGGCTGCCTGCC-3′ and reverseprimer: 5′-GCAGATCTTGCTCCCAGACCTGACC-3′) end-labeled with[γ-32P]ATP (New England Nuclear Corporation, Boston, MA) by T4polynucleotide kinase (Promega) in ULTRAhyb ultrasensitivehybridization buffer (Ambion). Free radio-labeled probe waswashed away with 2× SSC/0.1% SDS followed by 0.1× SSC/0.1% SDS.A final wash with 2× SSC was completed before wrapping themembrane in plastic and exposing to a phosphor screen and scanningon the molecular imager FX system (Bio-Rad Laboratories). Northernblot analysis was performed by isolating total RNA from RK13 cellsinfected with RacL11 EHV-1 at the indicated times using the RNA-Bee RNA isolation reagent (AMS Biotechnology (Europe) Ltd,Abingdon, U.K.) per the manufacturer's procedure (Chomczynskiand Sacchi, 1987). RNA samples were separated on a 6% denaturingurea-polyacrylamide gel. The above procedures were followed fortransferring and probing the RNA on the membranes. The probeused was a short fragment of the UL4 DNA (5′-TTATCGTT-TATTTTCTCGCTGGCGCTCTTTGGCCGAGGTTATTCCCCTAG-3′). West-ern blot analyses were performed using protein lysates fromRK13 cells infected with virus or transfected with expressionplasmids. The protein samples were separated on 10% SDSpolyacrylamide gels and transferred to nitrocellulose membranesusing the semi-dry electroblotter. Membranes were blocked with1-5% dry milk in PBS/0.3% Tween-20 (PBST). Membranes wererinsed with PBST and incubated with EHV-1 specific primaryantibodies, as indicated, followed by goat anti-rabbit or anti-mousesecondary antibody conjugated to alkaline phosphatase (Sigma,Saint Louis, MO). Protein-antibody complexes were visualizedusing the AP color reagent (Bio-Rad Laboratories) following themanufacturer's instructions.

Use of metabolic inhibitors

The procedure for virus infection with the use of metabolicinhibitors is described elsewhere (Gray et al., 1987). To detect

376 R.A. Charvat et al. / Virology 412 (2011) 366–377

immediate early transcripts, cells were incubated in the presence ofthe protein synthesis inhibitor cycloheximide (CHX, 100 μg/mL;Sigma) for 30 min prior to infection with EHV-1. Cells were infectedwith wild-type EHV-1 (MOI of 10) and maintained in the presence ofCHX for 4 h, at which time RNAwas isolated. To distinguish early fromlate transcripts, cells were treated with phosphonoacetic acid (PAA,100 μg/mL; Sigma) for 30 min prior to infection and then for 8 h afterinfection when RNA was isolated. The RNA was subjected to northernblot analysis as detailed above.

Growth kinetics, cell tropism and plaque morphology

Cells were seeded into 60 mm dishes to 80% confluency andinfected at an MOI of 5. Virus was harvested from the cells andsupernatant at the indicated times post-infection and serially dilutedto perform plaque assays on RK13 monolayers. Infected monolayerswere incubated with medium containing 1.5% methylcellulose, andplaques were enumerated after three days by fixing with 10%formalin and staining with 0.5% methylene blue (Perdue et al.,1974). Plaque morphology was examined using the ImageJ software(NIH, http://rsbweb.nih.gov/ij/).

Animal experiments and statistical analysis

Animal experiments were conducted as published previously(Osterrieder et al., 1996; von Einem et al., 2004). Groups (n=12) of 6week-old female CBA mice were intranasally infected with1.25×106 pfu of wild-type RacL11, UL4aa18stop, or ΔUL4R EHV-1 ormock-infected with sterile medium. Mice were weighed beforeinfection and every day post-infection for 9 days to monitor changesin body weight. Mice were sacrificed at days 2, 3, and 4 post-infection,and whole lungs were harvested.Whole lungs were also isolated fromanimals that succumbed to virus infection. Lung tissue was disruptedusing silica beads and BeadBeater (BioSpec Products, Inc., Bartlesville,OK), and virus was titered as described above. Statistical analysis wasperformed using the two-tailed Student's-t test to compare changes inbody weight as well as mortality rates.

Gene expression in cells infected with wild-type and UL4aa18stop EHV-1

RK13 cells were infected with wild-type or UL4aa18stop EHV-1 todetermine whether gene expression differs in the absence of theUL4P. RNA was isolated at 4, 8, and 12 hpi and examined by northernblot analysis using probes specific for the IE, TK and gK genes.

Acknowledgments

We thank Mrs. Suzanne Zavecz for excellent technical assistanceand the members of our laboratory for their helpful suggestions.These investigations were supported by research grant AI-22001from the National Institute of Allergy and Infectious Diseases andgrant P20-RR018724 from the National Center for ResearchResources of the National Institutes of Heath.

References

Ahn, B.C., Breitenbach, J.E., Kim, S.K., O'Callaghan, D.J., 2007. The equine herpesvirus-1IR3 gene that lies antisense to the sole immediate-early (IE) gene is trans-activatedby the IE protein, and is poorly expressed to a protein. Virology 363, 15–25.

Ahn, B.C., Zhang, Y., O'Callaghan, D.J., 2010. The equine herpesvirus-1 (EHV-1) IR3transcript downregulates expression of the IE gene and the absence of IR3 geneexpression alters EHV-1 biological properties and virulence. Virology 402,327–337.

Albrecht, R.A., Kim, S.K., Zhang, Y., Zhao, Y., O'Callaghan, D.J., 2004. The equineherpesvirus 1 EICP27 protein enhances gene expression via an interaction withTATA box-binding protein. Virology 324, 311–326.

Allen, G.P., Bryans, J.T., 1986. Molecular epizootiology, pathogenesis, and prophylaxis ofequine herpesvirus-1 infections. Prog. Vet. Microbiol. Immunol. 2, 78–144.

Baumann, R.P., Dauenhauer, S.A., Caughman, G.B., Staczek, J., O'Callaghan, D.J., 1984.Structure and genetic complexity of the genomes of herpesvirus defective-interfering particles associated with oncogenic transformation and persistentinfection. J. Virol. 50, 13–21.

Baumann, R.P., Staczek, J., O'Callaghan, D.J., 1986. Cloning and fine mapping the DNA ofequine herpesvirus type one defective interfering particles. Virology 153, 188–200.

Block, T., Jordan, R., Farkas, D.H., Hughes Jr., R.G., 1991. Inhibition of transient geneexpression with plasmids encoding herpes simplex virus type 1 UL55 and alphagenes. J. Gen. Virol. 72, 131–141.

Bowles, D.E., Holden, V.R., Zhao, Y., O'Callaghan, D.J., 1997. The ICP0 protein of equineherpesvirus 1 is an early protein that independently transactivates expression of allclasses of viral promoters. J. Virol. 71, 4904–4914.

Bowles, D.E., Kim, S.K., O'Callaghan, D.J., 2000. Characterization of the trans-activationproperties of equine herpesvirus 1 EICP0 protein. J. Virol. 74, 1200–1208.

Breitenbach, J.E., Ebner, P.D., O'Callaghan, D.J., 2009. The IR4 auxiliary regulatoryprotein expands the in vitro host range of equine herpesvirus 1 and is essential forpathogenesis in the murine model. Virology 383, 188–194.

Campbell, D.E., Kemp, M.C., Perdue, M.L., Randall, C.C., Gentry, G.A., 1976. Equineherpesvirus in vivo: cyclic production of a DNA density variant with repetitivesequences. Virology 69, 737–750.

Caughman, G.B., Staczek, J., O'Callaghan, D.J., 1985. Equine herpesvirus type 1 infectedcell polypeptides: evidence for immediate early/early/late regulation of viral geneexpression. Virology 145, 49–61.

Chen, M., Harty, R.N., Zhao, Y., Holden, V.R., O'Callaghan, D.J., 1996. Expression of anequine herpesvirus 1 ICP22/ICP27 hybrid protein encoded by defective interferingparticles associated with persistent infection. J. Virol. 70, 313–320.

Chen, M., Garko-Buczynski, K.A., Zhang, Y., O'Callaghan, D.J., 1999. The defectiveinterfering particles of equine herpesvirus 1 encode an ICP22/ICP27 hybrid proteinthat alters viral gene regulation. Virus Res. 59, 149–164.

Chomczynski, P., Sacchi, N., 1987. Single-step method of RNA isolation by acidguanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162,156–159.

Colle III, C.F., Tarbet, E.B., Grafton, W.D., Jennings, S.R., O'Callaghan, D.J., 1996. Equineherpesvirus-1 strain KyA, a candidate vaccine strain, reduces viral titers in micechallenged with a pathogenic strain, RacL. Virus Res. 43, 111–124.

Csellner, H., Walker, C., Wellington, J.E., McLure, L.E., Love, D.N., Whalley, J.M., 2000.EHV-1 glycoprotein D (EHV-1 gD) is required for virus entry and cell–cell fusion,and an EHV-1 gD deletion mutant induces a protective immune response in mice.Arch. Virol. 145, 2371–2385.

Dauenhauer, S.A., Robinson, R.A., O'Callaghan, D.J., 1982. Chronic production ofdefective-interfering particles by hamster embryo cultures of herpesviruspersistently infected and oncogenically transformed cells. J. Gen. Virol. 60, 1–14.

Ebner, P.D., O'Callaghan, D.J., 2006. Genetic complexity of EHV-1 defective interferingparticles and identification of novel IR4/UL5 hybrid proteins produced duringpersistent infection. Virus Genes 32, 313–320.

Ebner, P.D., Kim, S.K., O'Callaghan, D.J., 2008. Biological and genotypic properties ofdefective interfering particles of equine herpesvirus 1 that mediate persistentinfection. Virology 381, 98–105.

Elliott, G.D., O'Hare, P., 1995. Equine herpesvirus 1 gene 12, the functional homologue ofherpes simplex virus VP16, transactivates via octamer sequences in the equineherpesvirus IE gene promoter. Virology 213, 258–262.

Flowers, C.C., O'Callaghan, D.J., 1992. Equine herpesvirus 1 glycoprotein D: mapping ofthe transcript and a neutralization epitope. J. Virol. 66, 6451–6460.

Frampton Jr., A.R., Smith, P.M., Zhang, Y., Matsumura, T., Osterrieder, N., O'Callaghan, D.J., 2002. Contribution of gene products encoded within the unique short segment ofequine herpesvirus 1 to virulence in a murine model. Virus Res. 90, 287–301.

Frampton Jr., A.R., Smith, P.M., Zhang, Y., Grafton, W.D., Matsumura, T., Osterrieder, N.,O'Callaghan, D.J., 2004. Meningoencephalitis in mice infected with an equineherpesvirus 1 strain KyA recombinant expressing glycoprotein I and glycoprotein E.Virus Genes 29, 9–17.

Frampton Jr., A.R., Goins, W.F., Cohen, J.B., von Einem, J., Osterrieder, N., O'Callaghan, D.J., Glorioso, J.C., 2005. Equine herpesvirus 1 utilizes a novel herpesvirus entryreceptor. J. Virol. 79, 3169–3173.

Gray, W.L., Baumann, R.P., Robertson, A.T., Caughman, G.B., O'Callaghan, D.J., Staczek, J.,1987. Regulation of equine herpesvirus type 1 gene expression: characterization ofimmediate early, early, and late transcription. Virology 158, 79–87.

Gray, W.L., Yalamanchili, R., Raengsakulrach, B., Baumann, R.P., Staczek, J., O'Callaghan,D.J., 1989. Viral transcripts in cells infected with defective interfering particles ofequine herpesvirus type 1. Virology 172, 1–10.

Grundy, F.J., Baumann, R.P., O'Callaghan, D.J., 1989. DNA sequence and comparativeanalyses of the equine herpesvirus type 1 immediate early gene. Virology 172,223–236.

Harty, R.N., O'Callaghan, D.J., 1991. An early gene maps within and is 3' coterminal withthe immediate-early gene of equine herpesvirus 1. J. Virol. 65, 3829–3838.

Harty, R.N., Holden, V.R., O'Callaghan, D.J., 1993. Transcriptional and translationalanalyses of the UL2 gene of equine herpesvirus 1: a homolog of UL55 of herpessimplex virus type 1 that is maintained in the genome of defective interferingparticles. J. Virol. 67, 2255–2265.

Holden, V.R., Harty, R.N., Yalamanchili, R.R., O'Callaghan, D.J., 1992. The IR3 gene ofequine herpesvirus type 1: a unique gene regulated by sequences within the intronof the immediate-early gene. DNA Seq. 3, 143–152.

Holden, V.R., Zhao, Y., Thompson, Y., Caughman, G.B., Smith, R.H., O'Callaghan, D.J.,1995. Characterization of the regulatory function of the ICP22 protein of equineherpesvirus type 1. Virology 210, 273–282.

Huang, A.S., Baltimore, D., 1970. Defective viral particles and viral disease processes.Nature 226, 325–327.

377R.A. Charvat et al. / Virology 412 (2011) 366–377

Jang, H.K., Albrecht, R.A., Buczynski, K.A., Kim, S.K., Derbigny, W.A., O'Callaghan, D.J.,2001. Mapping the sequences that mediate interaction of the equine herpesvirus 1immediate-early protein and human TFIIB. J. Virol. 75, 10219–10230.

Kim, S.K., O'Callaghan, D.J., 2001. Molecular characterizations of the equine herpesvirus1 ETIF promoter region and translation initiation site. Virology 286, 237–247.

Kim, S.K., Holden, V.R., O'Callaghan, D.J., 1997. The ICP22 protein of equine herpesvirus1 cooperates with the IE protein to regulate viral gene expression. J. Virol. 71,1004–1012.

Kim, S.K., Bowles, D.E., O'Callaghan, D.J., 1999. The gamma2 late glycoprotein Kpromoter of equine herpesvirus 1 is differentially regulated by the IE and EICP0proteins. Virology 256, 173–179.

Kim, S.K., Ahn, B.C., Albrecht, R.A., O'Callaghan,D.J., 2006. The unique IR2 protein of equineherpesvirus 1 negatively regulates viral gene expression. J. Virol. 80, 5041–5049.

Lewis, J.B., Thompson, Y.G., Caughman, G.B., 1993. Transcriptional control of the equineherpesvirus 1 immediate early gene. Virology 197, 788–792.

Lewis, J.B., Thompson, Y.G., Jenkins, A.C., Caughman, G.B., 1995. Characterization andlocalization of the equine herpesvirus 1 major DNA binding protein. Virology 207,380–391.

Lewis, J.B., Thompson, Y.G., Feng, X., Holden, V.R., O'Callaghan, D.J., Caughman, G.B.,1997. Structural and antigenic identification of the ORF12 protein (alpha TIF) ofequine herpesvirus 1. Virology 230, 369–375.

Matsumura, T., O'Callaghan, D.J., Kondo, T., Kamada, M., 1996. Lack of virulence of themurine fibroblast adapted strain, Kentucky A (KyA), of equine herpesvirus type 1(EHV-1) in young horses. Vet. Microbiol. 48, 353–365.

Mettenleiter, T.C., Keil, G.M., Fuchs, W., 2008. Molecular biology of animalherpesviruses. In: Mettenleiter, T.C., Sobrino, F. (Eds.), Animal Viruses: MolecularBiology. Caister Academic Press, Norfolk, UK, pp. 375–455.

Misra, V., Bratanich, A.C., Carpenter, D., O'Hare, P., 1994. Protein and DNA elementsinvolved in transactivation of the promoter of the bovine herpesvirus (BHV) 1 IE-1transcription unit by the BHV alpha gene trans-inducing factor. J. Virol. 68,4898–4909.

Moriuchi, H., Moriuchi, M., Cohen, J.I., 1995. Proteins and cis-acting elements associatedwith transactivation of the varicella-zoster virus (VZV) immediate-early gene 62promoter by VZV open reading frame 10 protein. J. Virol. 69, 4693–4701.

Muramatsu, M., Smetana, K., Busch, H., 1963. Quantitative aspects of isolation ofnucleoli of theWalker carcinosarcoma and liver of the rat. Cancer Res. 23, 510–518.

O'Callaghan, D.J., Osterrieder, N., 2008. Herpesviruses of horses, Third ed. In:Mahy, B.W.J., VanRegenmortel, M.H.V. (Eds.), Encyclopedia of Virology, Vol. 2. Elsevier Ltd.,Oxford, pp. 411–420.

O'Callaghan, D.J., Randall, C.C., 1976. Molecular anatomy of herpesviruses: recentstudies. Prog. Med. Virol. 22, 152–210.

Osterrieder, N., Holden, V.R., Brandmuller, C., Neubauer, A., Kaaden, O.R., O'Callaghan, D.J., 1996. The equine herpesvirus 1 IR6 protein is nonessential for virus growth invitro and modified by serial virus passage in cell culture. Virology 217, 442–451.

Perdue, M.L., Kemp, M.C., Randall, C.C., O'Callaghan, D.J., 1974. Studies of the molecularanatomy of the L-M cell strain of equine herpes virus type 1: proteins of thenucleocapsid and intact virion. Virology 59, 201–216.

Poffenberger, K.L., Raichlen, P.E., Herman, R.C., 1993. In vitro characterization of aherpes simplex virus type 1 ICP22 deletion mutant. Virus Genes 7, 171–186.

Post, L.E., Roizman, B., 1981. A generalized technique for deletion of specific genes inlarge genomes: alpha gene 22 of herpes simplex virus 1 is not essential for growth.Cell 25, 227–232.

Purewal, A.S., Allsopp, R., Riggio, M., Telford, E.A., Azam, S., Davison, A.J., Edington,N., 1994. Equid herpesviruses 1 and 4 encode functional homologs of theherpes simplex virus type 1 virion transactivator protein, VP16. Virology 198,385–389.

Rudolph, J., O'Callaghan, D.J., Osterrieder, N., 2002. Cloning of the genomes of equineherpesvirus type 1 (EHV-1) strains KyA and racL11 as bacterial artificialchromosomes (BAC). J. Vet. Med. B Infect. Dis. Vet. Public Health 49, 31–36.

Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual,2nd ed. Cold Springs Harbor Laboratory Press, Cold Springs Harbor, NY.

Smith, R.H., Caughman, G.B., O'Callaghan, D.J., 1992. Characterization of the regulatoryfunctions of the equine herpesvirus 1 immediate-early gene product. J. Virol. 66,936–945.

Smith, R.H., Zhao, Y., O'Callaghan, D.J., 1994. The equine herpesvirus type 1immediate-early gene product contains an acidic transcriptional activationdomain. Virology 202, 760–770.

Smith, P.M., Zhang, Y., Grafton, W.D., Jennings, S.R., O'Callaghan, D.J., 2000. Severemurine lung immunopathology elicited by the pathogenic equine herpesvirus 1strain RacL11 correlates with early production of macrophage inflammatoryproteins 1alpha, 1beta, and 2 and tumor necrosis factor alpha. J. Virol. 74,10034–10040.

Smith, P.M., Kahan, S.M., Rorex, C.B., von Einem, J., Osterrieder, N., O'Callaghan, D.J.,2005. Expression of the full-length form of gp2 of equine herpesvirus 1 (EHV-1)completely restores respiratory virulence to the attenuated EHV-1 strain KyA inCBA mice. J. Virol. 79, 5105–5115.

von Einem, J., Wellington, J., Whalley, J.M., Osterrieder, K., O'Callaghan, D.J., Osterrieder,N., 2004. The truncated form of glycoprotein gp2 of equine herpesvirus 1 (EHV-1)vaccine strain KyA is not functionally equivalent to full-length gp2 encoded byEHV-1 wild-type strain RacL11. J. Virol. 78, 3003–3013.

von Einem, J., Schumacher, D., O'Callaghan, D.J., Osterrieder, N., 2006. The alpha-TIF(VP16) homologue (ETIF) of equine herpesvirus 1 is essential for secondaryenvelopment and virus egress. J. Virol. 80, 2609–2620.

Warming, S., Costantino, N., Court, D.L., Jenkins, N.A., Copeland, N.G., 2005. Simple andhighly efficient BAC recombineering using galK selection. Nucleic Acids Res. 33,e36.

Yamada, H., Jiang, Y.M., Oshima, S., Daikoku, T., Yamashita, Y., Tsurumi, T., Nishiyama, Y.,1998. Characterization of the UL55 gene product of herpes simplex virus type 2. J.Gen. Virol. 79, 1989–1995.

Zhang, Z., Rowe, J., Wang, W., Sommer, M., Arvin, A., Moffat, J., Zhu, H., 2007. Geneticanalysis of varicella-zoster virus ORF0 to ORF4 by use of a novel luciferase bacterialartificial chromosome system. J. Virol. 81, 9024–9033.

Zhao, Y., Holden, V.R., Smith, R.H., O'Callaghan, D.J., 1995. Regulatory function of theequine herpesvirus 1 ICP27 gene product. J. Virol. 69, 2786–2793.